Can AI explain itself? In other words, if I take an AI model and ask it questions about any another AI model, I would expect it to answer my questions. But it turns out that the model simply doesn't know about any other models.
Can a simple workaround for this problem lies with Retrieval Augmented Generation or RAG in short? This repo implements a RAG pipeline using LangFlow and StreamLit to develop an App called, "AI Explans AI".
Checkout the below YouTube video which walks through the app development.
- Clone the repo
git clone git@github.com:ai-bites/simple-rag-app.git
- Its important that the vwe are in
python 3.10
to get the installation right. Then we just needlangflow
streamlit
andlangchain-community
conda create -n rag-app python=3.10
conda activate rag-app
pip install langflow
pip install streamlit
pip install langchain-community
To run the app directly, just run streamlit on app.py
.
streamlit run app.py
If you just want to run langflow, run langflow and import the json
file
langflow run